Diatoms are a widespread group of microalgae and can be found in the oceans, freshwater, and in soils. Biomass from diatom culture (whether naturally or commercially produced) contains a number of products of commercial interest including lipids, fatty acids (particularly highly unsaturated fatty acids—HUFAs), amino acids, pigments and complex natural products of pharmacological interest. The biomass itself also has dietary application (particularly in aquaculture), and the treatment of water contaminated by phosphorus and nitrogen in aquaculture effluent, or heavy metal (bioremediation). A further application in development is the use of silicon derived from frustules (the cell wall or external layer of diatoms) in nanotechnology. However, the commercial relevance of products from diatoms will depend on the cost of their production.
Industrial fermentation is a costly process in terms of capital equipment, nutrients, and energy, and is generally only justified when a relatively high value product is being produced in large quantities.
Different modes of culturing or fermentation are possible. The simplest mode of fermentation and the one that is almost exclusively used in industrial processes is batch fermentation. In batch fermentation, cells are inoculated in nutrient medium, grown for a period of time and then harvested. Fed-batch fermentation is similar to batch but differs in that concentrated nutrients are supplied to the culture during the growth period. Continuous fermentation involves the continuous harvest of culture comprising biomass and a nutrient solution from the fermentation vessel and its replacement with fresh nutrient solution. The rate of harvest in continuous fermentation is chosen so that the density of the cells in culture remains constant. Semi-continuous fermentation is similar to continuous fermentation except that harvests are periodic rather than continuous. Perfusion fermentation involves the continuous harvest of a culture medium which contains the product of interest, while the cells producing the product are retained in the culture.
Where the product of interest is produced within the cells, the cells are generally cultivated to the highest biomass densities possible to get the most efficient volumetric productivity (i.e., amount of product produced per volume of fermentation medium per unit time), thereby minimizing the cost of production of the product of interest. However, at high biomass densities there can be difficulties in providing sufficient oxygen, nutrients and, where applicable, light. For example, in a batch or fed batch fermentation, microbial cultures are generally in a stationary phase at the point of harvest. It is usual for the medium or feeding regime in these types of fermentations to result in the culture being severely limited for one or more nutrients, often as a means of inducing the formation of secondary metabolites (which are often the products of commercial interest).
It was previously known that photosynthetic organisms (e.g., microalgae) can be grown phototrophically under continuous or semi-continuous culture conditions. Under these conditions light is used as the energy source, rather than reduced carbon. Many authors (e.g. Richardson et al. (1969) Applied Microbiology 18:245-250; Droop (1974) J. Mar. Biol. Ass. U.K. 54:825-855; Laing (1991) Lab.Leafl.MAFF Direct.Fish.Res., Lowestoft, (67) 31 pp) have disclosed means of growing algae in continuous or semi-continuous phototrophic culture (turbidostats). In such cultures, cells are generally limited by the amount of light that is available and so culture densities are low and thus volumetric productivity is very low. Advances in photobioreactor technology have overcome the light limitation problem to a certain extent by narrowing the optical plane and allowing cultures to reach higher biomass concentrations (Zou et al. (2000) Eur. J. Phycol. 35:127-133). However, this requires an increase in surface area to volume ratio of the culture and significantly increases the capital cost structure of the reactors. Furthermore, at these concentrations growth is slow (light limited) and thus volumetric productivity is again low. These photobioreactor systems also have the disadvantage that they would be uneconomic to construct for volumes large enough to supply industrial demand and even the best commercially scalable photobioreactor designs still produce biomass densities only about one-tenth that capable in a fermentor. Such photobioreactors are therefore generally considered to be just a laboratory tool for study of the growth of photosynthetic organisms.
Pharmaceuticals, medical foods and nutritional supplement containing highly unsaturated fatty acids (HUFAs) are currently used to treat hundreds of thousands, and potentially will soon be used to treat, millions of patients. Eicosapentaenoic acid (EPA) is a HUFA used as an active metabolite in drug substances. Docosahexaenoic acid (DHA) also has great potential for use in the pharmaceutical, medical food and nutritional supplement industry. An example of their use is the treatment or prophylactic treatment of cardiovascular disease.
HUFAs cannot be chemically synthesized de novo economically, therefore must be extracted from a biological source. Pharmaceutical manufacturers currently rely on fish as the source of HUFAs for production of drug substances. Exclusive reliance on fish oil for such purposes, however, carries a number of serious risks to pharmaceutical manufacturers and drug companies as well as potentially to patients receiving such medications. Such risks include, but are not limited to, those associated with potential supply shortages, which may be financially devastating to drug companies as well as negatively affect patients who rely on medications for their well-being. Fish oil itself is not a specific reference material since its composition differs dramatically between different fish species. Even within a single fish species the composition varies from location to location and it even varies at different times of the year at a single location. Such variability in starting material makes manufacturing a final pharmaceutical product very difficult. Furthermore there is a growing concern over the toxic pollutants such as poly chlorinated biphenyls (PCBs), dioxin, and methyl mercury that are showing up in fish oils. There are also widespread concerns over the long term sustainability of wild fish stocks and aquaculture.
There is an acute need therefore, for alternative sources of HUFAs to fish oil which are consistent, reliable and therefore amenable to commercial and sustainable production methods. Industrial fermentation is one possible alternative for commercial production of HUFAs. However, there are a number of problems with using industrial fermentation on a commercial scale.
A number of authors have disclosed continuous culture of microorganisms for the production of lipids (for example Hall and Ratledge (1977) App. Env. Microbiol 33:577-584; Gill et al. (1977) App. Env. Microbiol. 33231-239; Ykema et al. (1988) Appl. Microbiol. Biotechnol. 29:211-218; Brown et al. (1989) J. Ferm. Bioeng. 68:344-352; Kendrick and Ratledge (1992) Appl. Microbiol. Biotechnol. 37:18-22; Papanikolaou and Aggelis (2002) Bioresouce Technology 82:43-49). Whilst some of these do disclose high volumetric biomass productivity, none of the organisms disclosed produce any highly unsaturated fatty acids (HUFAs; fatty acids with 20 or more carbons and 4 or more double bonds).
Other types of organisms, in particular microalgae of the genus Crypthecodinium and marine fungi in the family Thraustochytriales (including Schizochytrium species, Thraustochytrium species and Ulkenia species) may be more uniquely suited to production of HUFAs, and particularly the omega-3 DHA. Whilst these organisms produce the greatest amounts of DHA under conditions of nitrogen limitation and are thus well suited to the fed batch cultures used in industrial processes, some authors have disclosed a continuous heterotrophic culture of these species for the production of DHA. However, most of these continuous cultures were only demonstrated in the laboratory and were at low production rates, which are not commercially viable.
Ganuza and Izquierdo ((2007) Appl. Microbiol. Biotechnol. 76:985-990) disclose continuous culture of Schizochytrium G13/2S for the production of DHA. Their maximum biomass productivity occurred at a dilution of 0.04 per hour and a dry weight of 7.7 g/L giving a volumetric productivity of only 7.4 g/L/day.
Ethier et al. ((2011) Bioresource Technology 102:88-93) disclose continuous culture of Schizochytrium limacinum for producing DHA. Their highest biomass productivity was only 3.88 g/L/day.
Pleisner and Errikson (http://www.marbio.sdu.dk/uploads/MarBioShell/Pleissner %20-%20VejleNuthetalPoster.pdf accessed 8 Feb. 2013), disclose continuous culture of Crypthecodinium cohnii for producing DHA, but only achieve a biomass productivity of 12 g/L/day. Further, at this volumetric productivity the HUFA content of the biomass was only 1.67%, making the final extraction and processing problematic.
In contrast, Wümpelmann in WO2005/021735 discloses continuous culture of Schizochytrium limnaceum for the production of DHA at high culture densities and through this achieves biomass productivities in excess of 100 g/L/day. The author discloses methods that are particularly suitable to producing DHA from this organism, indicating for instance that dissolved oxygen tension should be maintained at low levels, but does not provide sufficient guidance about whether such conditions can even be used for other types of microorganisms or, if so, how such conditions should be varied to allow these sorts of productivities for other organisms with completely different environmental and nutritional requirements. The methods disclosed by the author are also not well suited for production at an industrially relevant scale. Indeed, the examples disclosed are only carried out in 2 L fermenters in the laboratory and the cost of using just one of the media constituents (e.g., casamino acids) as a nitrogen source would be prohibitively expensive at scales of 100 L or more.
Other types of organisms, in particular the diatoms, may be more suited for the production of the omega-3 EPA, especially where it is advantageous for the EPA to be produced with relatively low amounts of DHA present.
Wen and Chen disclose the heterotrophic culture of the diatom Nitzschia laevis for the production of the omega-3 fatty acid EPA. However, in true continuous cultures their maximal biomass productivity was only 2.8 g/L/day ((2002) Biotechnol. Prog. 18:21-28), whilst with the addition of perfusion (wherein cells and media are separated, cells are returned to the fermenter and additional fresh media is added), this was raised to 6.75 g/L/day ((2001) Appl Microbiol Biotechnol 57:316-322).
Griffiths et al. (WO2011/155852) disclosed the heterotrophic culture of the diatom Nitzschia laevis both in perfusion-aided continuous culture and in true continuous culture. Whilst the authors do no not disclose the volumetric productivities of the cultures, culture dry weights were around 10 g/L or lower and in perfusion-aided continuous culture the total volume removed from the culture was one fermenter volume per day. Even if all this volume were harvest and none were perfusion, the maximum biomass productivity would be significantly higher than described by Wen and Chen, but would still only be around 10 g/L/day.